Currently available AR antagonists such as bicalutamide have limited activity in castration-resistant prostate cancer (CRPC). Mechanistic studies have shown that the bicalutamide liganded AR can accumulate in the nucleus and bind to chromatin, but that it does not effectively recruit coactivator proteins and instead can recruit the corepressor proteins NCoR and SMRT. This ability to mediate chromatin binding may be an Achilles Heal of these agents, as it could facilitate their acquisition of weak or possibly strong agonist activity in CRPC. In contrast to these currently available antagonists, MDV3100 does not enhance nuclear localization or simulate chromatin binding, and recently reported phase III trial results show that MDV3100 can prolong survival in CRPC. These findings likely will lead to FDA approval of MDV3100 in the near future, arming medical oncologists with two new agents (abiraterone and MDV3100) that can effectively suppress AR activity in CRPC. Therefore, it is now critical to understand in detail mechanisms of action for MDV3100 and related AR antagonists under development, to determine how they might best be used in conjunction with abiraterone, and to identify mechanisms of intrinsic and acquired resistance. Our overall objectives in this proposal are to understand how MDV3100 and functionally related AR antagonists interfere with AR binding to chromatin, and to identify mechanisms of resistance. Our recent studies have shown that phosphorylation of serine 81 (S81) in the AR N-terminal domain is critical for stable binding of AR to chromatin. Moreover, our data indicate that the primary mechanism of action of MDV3100 is to prevent chromatin binding, and that MDV3100 interferes with a S81 dependent step in AR chromatin binding. Based on these findings, our general approach is to determine the molecular basis for S81 dependent AR binding to chromatin (Aim 1) and to determine how MDV3100 and functionally related antagonists interfere with this binding (Aim 2). We will then evaluate candidate mechanisms of MDV3100 resistance in cell line models (Aim 2), and in xenograft models and clinical samples (Aim 3).